Machine Learning Formulation Network for Accelerated Design

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Modern product formulation is complex and inefficient due to dispersed institutional knowledge, leading to formulation data gaps, anomalies, and conflicts, which hinder the accuracy and quality of product formulations without a single source of truth.

Innovation Solution

A computer-implemented method using a network of distributed computing systems to generate unsupervised and supervised formulation network models, identifying dependency connections and outcome-contributory values for design variables, and rendering these models graphically to provide intelligent formulation recommendations.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If traditional manual formulation methods are used, then formulators have flexibility in experimentation, but formulation development is slow and time-consuming

Engineering Contradiction:

Improveformulation development speedVSAvoidtime for real-world experiments

Core Design Contradiction:

ProductivityVSLoss of time

Solution Approach 1:

The system performs preliminary computational formulation development by generating multiple candidate formulations through machine learning models before physical experimentation. The formulation network model predicts outcomes and identifies promising candidates in advance, allowing formulators to focus only on validating the most promising options through minimal real-world experiments.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system creates virtual copies of formulation experiments through computational modeling. By simulating formulation outcomes in a digital environment using the formulation network model, the system can evaluate multiple scenarios without requiring physical experimentation for each one, thereby reducing the time and resources needed for real-world testing.

Inventive Principle:

Principle #26Copying

2Loss of information

If formulation data is dispersed across multiple sources, then comprehensive knowledge is available, but data gaps, anomalies, and conflicts arise

Engineering Contradiction:

Improveformulation data completenessVSAvoidformulation data accuracy

Core Design Contradiction:

Loss of informationVSReliability

Solution Approach 1:

The system merges multiple dispersed formulation data sources into a unified formulation network model. By integrating data from various sources including literature, patents, and experimental results into a single coherent model structure, the system eliminates data gaps and resolves conflicts by establishing consistent relationships between formulation components and outcomes.

Inventive Principle:

Principle #5Merging (Combining)

Solution Approach 2:

The formulation network model incorporates feedback mechanisms that continuously validate and refine formulation predictions against actual experimental outcomes. This feedback loop allows the system to detect and correct data anomalies, ensuring that the model maintains high accuracy and reliability while comprehensively representing the formulation knowledge base.

Inventive Principle:

Principle #23Feedback

3Adaptability or versatility

If expert formulation knowledge is not memorialized systematically, then flexibility in application remains, but knowledge transfer to new formulations is difficult

Engineering Contradiction:

Improveformulator expertise flexibilityVSAvoidknowledge transfer efficiency

Core Design Contradiction:

Adaptability or versatilityVSEase of manufacture

Solution Approach 1:

The formulation network model serves as a universal knowledge base that captures expert formulation insights in a standardized, machine-readable format. This universal model can be applied across different formulation problems and product types, enabling efficient knowledge transfer while maintaining the flexibility to adapt to specific application requirements through configurable model parameters and constraints.

Inventive Principle:

Principle #6Universality (Multi-functionality)

Data Source

PatentUS11783103B2Systems and methods for an accelerating product formulation creation via implementing a machine learning-derived graphical formulation network model

Publication Date: 2023.10.10 TURING LABS INC

AI summary

A method and system for implementing one or more machine learning models for accelerating formulation design for a target product that includes converting an unsupervised formulation network model to a supervised formulation network model, deriving an outcome-contributory value for each of a plurality of distinct design variables of the supervised formulation network, identifying a dependency connection between each of a plurality of distinct pairs of distinct design variables, computing a strength of connection metric value for each of the plurality of distinct pairs of distinct design variables; and generating, via a graphical user interface, a graphical rendering of the supervised formulation model that may be manipulated to accelerate for design of a proposed formulation for a target physical product.